I’ve moved from Germany to France over the past few days for another presentation. Currently, I’m in Lyon and spent some off time today exploring the old city. This morning included, I took side trip to a shop in the old city where they still have a couple of working Jacquard looms, which employ a punched card system for determining the pattern that would be woven into cloth. Yes, this “programming” system was an early forerunner for early computer systems that also used punched cards to store programs and input/output data.
A Jacquard loom actually has a very complex system of punched cards that tell the system of treadles which threads to raise. I’m not an expert on weaving but my first wife had a large loom in our first home and I learned enough about how weaving works to see the relationship to computing.
The essence of digital processing is binary counting. A specific column on a specific row of a punch card can either be punched or not. This is a two state system and forms the basis of a stored program computer. In the world of digital audio, we use binary counting to store the amplitude of an analog waveform at specific moments or samples in time.
There’s a great deal of complexity to encoding digital audio but Pulse Code Modulation is a well-established system, which can capture and reproduces very high fidelity audio. In spite of repeated and incorrect claims by those advocating (and benefiting) from other methods including Pulse Width Modulation, PCM remains the primary production digitization process used in the world of audio and the best one.
I read a recent piece by Andreas Koch, a principal of Playback Designs, called, “The Subliminal Impact of Sound” at the Primephonic website. He suggests that DSD is somehow more “analog” sounding than PCM (as if analog sounding is a desirable goal!!). Aren’t we looking for accurate reproduction of sound instead?
You can read his article for yourself, but my favorite line is:
“As explained in an earlier article the frequency range of DSD does not have any cliffs, has a flat area for the classic audio range and then decays slowly at higher frequencies. Because of its very high sample rate the usable frequency band goes up to 1.4MHz for the slowest sample rate of 2.8Mhz.”
His claim that a DSD recording has “usable” frequency band that extends “up to 1.4 MHz” is pure baloney! What he has conveniently left out of his comparison is the fact that all of the noise present in the audio band (20-20 kHz) has been shifted up beyond 20 kHz and makes heavy filtering necessary on playback to remove the excessive noise in the ultrasonic range (you don’t want the ultrasonic noise making it to your equipment or speakers. A DSD 64 file doesn’t exceed the usable frequency range of a CD by very much…certainly not anywhere close to a properly done 96 kHz/24-bit PCM file. How easy it is to stray from the facts when you have a vested interest in advancing one particular over another.
High-resolution PCM digital encoding is a vast improvement in recording and reproducing music over any analog predecessor. It is more accurate, contains less noise, has less distortion, and is able to capture every dynamic nuance better than vinyl LPs, analog tape, CDs, and DSD files. Many will try to challenge this but the facts are the facts.
BTW The image used today is one I took of a miniature scene in a shop in Lyon, France. It contains a model of a weaving room with a Jacquard loom. Look carefully, it spooky cool.